Proposed is a study is to assess the relations between neighborhood characteristics (e.g., neighborhood disadvantage, collective efficacy, residential segregation) and drug use behavior among residents of New York City while controlling for relevant individual-level factors. A secondary purpose of this study is to determine the relation between key neighborhood characteristics and the medical consequences of drug use, particularly fatal drug overdose. Recent work has shown that neighborhood characteristics are associated with the likelihood of drug use, abuse, and dependence and that research focusing exclusively on individual patterns of drug use is insufficient to fully explain inter-individual drug use behavior. Here we propose a systematic collection of data to characterize neighborhoods in a large metropolitan area and to test key hypotheses that may explain inter-individual variability in drug use behavior and its consequences. Understanding how neighborhood factors affect drug use risk behavior can guide interventions that aim to mitigate drug use and its consequences, to create a social environment that better facilitates users to seek treatment, and to improve overall population health. In order to achieve the aims of this study we will use data from three sources: (i) We will collect data on individual drug use behavior (focusing on cigarette, alcohol, and marijuana use) through a random digit-dial phone survey of 5,000 New York City residents, geo-coding and mapping each respondent to their neighborhood of residence; (ii) we will compile a contextual database characterizing New York City neighborhoods using information collected in the survey and using archival data (collected from the New York City Department of Health and Mental Hygiene, Department of City Planning, Police Department, US Census and other sources); and (iii) we will combine this data with data from a NIDA funded study of drug overdose over time (DA 06534; principal investigator: Tardiff, a consultant on this project). We will test key hypotheses about the relations between neighborhood characteristics and drug use, abuse, and dependence using multi-level hierarchical models. We have recently completed several studies of New York City neighborhoods that serve as preliminary exploration of the key hypotheses of this study. This, coupled with our studies involving telephone interviews with 6,000+ New Yorkers is assurance that the aims of this study can be achieved.